Rapid Modeling Of Simple Ceramic Model

Ceramics for daily use are very active in daily life,and most of the shapes are simple and not complicated,such as bowls,dishes,plates,altar,vats,bottles,tea sets,etc.,are very regular or regular simple polygons(bodies).In the ceramic customization industry,customers often find the porcelain they like through various channels and customize the porcelain with different patterns of the same type.It is inevitable that there will be various deviations in the communication with craftsmen through oral description and photos.However,after obtaining the simple ceramic shape through photos or pictures,you can freely edit the peribody pattern,so as to expand the selection range of customers and improve user experience.In the field of image segmentation,the traditional segmentation algorithm will be unable to automatically deduct the target image with certain probability and rely on the manual auxiliary operation,resulting in the cost of manpower in the image preprocessing.With a certain probability,the traditional segmentation algorithm will be unable to automatically deduct the target image and rely on the manual auxiliary operation,which will result in the labor cost in the preprocessing of the image.On the other hand,under the condition of ensuring the accuracy,the general image semantic segmentation cannot be applied to the ceramic customization industry very well.In the first half of this paper,the existing segmentation methods based on deep learning are studied to achieve the automatic segmentation of target ceramic images and the effect of target edge refinement and optimization.After extracting the target ceramic from the image and separating it,we processed the foreground image into a three-dimensional model.In the field of 3d reconstruction,the research work mainly includes feature detection and matching,space constraint matrix calculation,camera calibration and 3d reconstruction system.The mature methods of 3d reconstruction algorithm include modeling based on single view,modeling based on double(multiple)view,modeling based on single(dual)view video and so on.Considering the actual situation of the ceramic customization industry,this paper focuses on the single-view modeling method,among which the representative method is SFS(Shape from Shading),that is,light and shade restoration method.With the development of computer vision theory and the proposal of various 3d reconstruction strategies,leading scholars and researchers in many countries have developed and produced perfect 3d reconstruction systems,and related products havebeen active in many fields such as medical treatment,industry,military,entertainment,virtual reality and so on.To sum up,this paper mainly studies the process method of simple ceramic vessel image segmentation and output as 3D model for ceramics.